Consumer Judgments, Decisions and Framing Dynamics: an Informational Viewpoint

Narasimhan Srinivasan, University of Connecticut
[ to cite ]:
Narasimhan Srinivasan (1993) ,"Consumer Judgments, Decisions and Framing Dynamics: an Informational Viewpoint", in NA - Advances in Consumer Research Volume 20, eds. Leigh McAlister and Michael L. Rothschild, Provo, UT : Association for Consumer Research, Pages: 288-290.

Advances in Consumer Research Volume 20, 1993      Pages 288-290

CONSUMER JUDGMENTS, DECISIONS AND FRAMING DYNAMICS: AN INFORMATIONAL VIEWPOINT

Narasimhan Srinivasan, University of Connecticut

The use of knowledge can be categorized along several dimensions:

(A) Composition of the decision making unit

(a) Individual, or (b) Joint decision making;

(B) Product Category

(a) Durables, or (b) Non-durables;

(C) Context or situation

(a) Memory based, or (b) Stimulus based; and

(D) Type of decision

(a) Judgment, or (b) Choice.

Conceptually, this embraces the well known person/product/context dimensions and considers the complex nature of decision making. In addition, it is also meant to include possible additional interactions. The three papers to be discussed form a subset of the various possibilities that one may construct using just the four dimensions mentioned, which provides the overarching framework. (For example, the design for Paper 1 may be denoted as AbBaCaDb). Brief discussions of the three papers follow:

FRAMING DYNAMICS: MEASUREMENT ISSUES AND PERSPECTIVES

This paper is rich in conceptualizing framing dynamics. Puto (1987) detailed a behavioral model of the buying decision framing process using a hypothetical buying scenario using 372 professional buyers. This study of 230 home buyers, representing a non-professional group of buyers, is a good extension.

The explicit recognition of process dynamism in purchase decisions is a highlight of this paper. Since a considerable time period passes between the onset of the "problem recognition" stage and the "final" purchase, the consumer goes through an intensive cognitive effort for important purchases (such as houses), and the presence of several environmental cues contributes toward a complex purchasing situation informationally.

It appears that the field work involved interviewing both the male and female heads of the household. Joint decision making is a very complex and exciting area of work to apply decision frames and I wish more attention to this issue is forthcoming in the future.

The authors point out that carry over effects, latency effect and learning effects present measurement problems in process data. Unfortunately, these problems have neither been measured nor addressed in the empirical part of the paper.

Six dynamic frames have been proposed, with the first five mentioning "reference prices" (plural emphasized). Maybe, the operationalization is not complete in Table 1. Table 1 does not detail how multiple figures (reference prices) were obtained for the decision frames. DF1 measures the "price at which the (previous) house was sold" or "monthly rental charge." Similarly, DF5 asks about "any other house(s) you seriously considered buying..." What if the response is "0" or "1?" A clearer demonstration of the correspondence between the conceptualization and operationalization would be helpful.

Some other specific issues that may be clarified in future work include:

- Usefulness of a global measure of variance: Are the home buyers homogenous in any meaningful respect? (particularly when the authors report a "wide range of house prices and local market size differences.")

- Segmentation of the market and different framing mechanisms for each segment might prove beneficial.

- The measure of "pattern" (Table 2) talks about the coefficient of variance as mean/range for the current frame. Shouldn't range be manipulated to use it as a measure of standard deviation?

- Propositions are not really tested at all. Rigorous testing of the propositions with clear demarcation of the decision frames would be a contribution to the literature.

- The supply side of framing is captured. What about the demand side perspective? Eagerness and willingness to expose oneself to more and more information and the real estate agent trying to "close the sale?"

THE IMPACT OF DIRECTION-OF-COMPARISON ON THE FORMATION OF PREFERENCES

This is a very interesting paper exploring the intricacies of a seemingly easy comparison of a pair of alternatives. The positive aspects of the paper include (1) the search for the boundary conditions for the direction-of-comparison effects and (2) testing under two contexts: stimulus-based judgment and memory-based judgment.

It appears that this study is a pre-cursor to a more elaborate design having three factors at two levels each: positive features (unique/shared) X negative features (unique/shared) and context (memory based and stimulus based).

A methodological concern I have is how many attributes, whether positive or negative does one include in a study. Perhaps the debate on the information overload about a decade earlier can provide some guidance. Providing just a single minute to memorize eight attributes for a brand seems to be a difficult and unrealistic task, particularly for a product like breakfast cereal. What is the motivation to assimilate all the available information? Did subjects use only part of the information provided to them? In the supermarket, consumers may take just a few seconds but don't register a lot of information in memory at least, according to Dickson and Sawyer (1990).

I am intrigued by the authors' statement that pretesting established that unique features of one brand were equal in desirability to the unique features of the other brand. I wish more details were made available. Even when brands share most all the significant technical specifications (e.g. Coke and Pepsi?), preference need not be based on that which is shared. How can unique features be taken to be equivalent, when they are different? Wouldn't it be useful to distinguish between attributes which are salient and those which are determinant?. Is desirability taken to be synonymous with both? When there is a partial sharing of attributes, there can be additional complications which are worth looking at.

It is easy to visualize someone replacing an existing brand either with the same brand (loyal) or a different brand (non-loyal). In such cases, the existing brand is the subject of comparison and the new brand is the referent of comparison. In the present set-up, how do you account for primacy and recency effects? Moreover, aren't such effects moderated by the level of involvement and the time between the arguments, according to the persuasion literature? What is the motivation to process more or less information?

FIGURE

SOME DIMENSIONS OF INFORMATION

If I may abstract the preference ratings roughly:

TABLE

Clearly, there appears to be a reversal between the two contexts. Is this caused by the sharing of the features? It appears to be worthy of some follow-up work. Taking pair-wise contrasts and drawing some more inferences appears worthwhile. Which of them are really different from the mid-point, which is no preference, indicative of low involvement? What is the power of the tests?

Some additional points to ponder about:

- We have been talking about judgment, not choice. What happens if subjects are asked to choose?

- Though only a pair of alternatives have been compared, what happens when the evoked set contains more than 2 brands?

EFFECT OF MULTIPLE MEASUREMENT OPERATIONS ON CONSUMER JUDGMENT: MEASUREMENT RELIABILITY OR REACTIVITY?

I enjoyed reading this paper on the assessment of measurement effects on judgment strength. However, I think the title is misleading: it ought to have said "Repeated Measures," instead of "Multiple Measures."

There are several positive features in this study. As any good study does, this paper raises a lot of interesting questions:

- Is judgment distinguishable from attitude? Generalized feeling, comparison alternative, indication of purchase intent, etc.?

- 56 brands are flashed on the screen three times successively. Why is it surprising that people get quicker in pressing buttons, though fatigue might also have set in?

- How is learning accounted for? History effects? Experimentation effects?

- How about expert judgment? Automaticity of responses? (Alba and Hutchinson 1987)

- Judgments may exist when not measured, or constructed. Which judgments will be effected and to what degree by multiple measurements?

- Attitudes may reach a plateau beyond which measurement effect is not a problem, perhaps. When will this point be reached?

Purchase frequency estimates are used to identify the "non-loyal" group. Why is it not possible that one hates some brand and hence never buys this brand; according to your computation, this person will have non-crystallized attitudes because of the non-loyal categorization, though this is clearly wrong. Attitude crystallization should not be confused with valence.

I am curious about what happened to the huge number of brands which are between your "loyal" and "non loyal" groups? Wouldn't it be interesting to know what happens when opinions are not polarized?

Though both the "loyal" and "non loyal" groups show declining response times, the latency for the non-loyal group reaches 104 centiseconds in the third exposure, which is not very different from the 106 centiseconds for the first exposure for loyals. Wonder if there is a merging? What does it imply for the strength of judgment? Are we really measuring for uncrystallized attitudes and their malleability?

I think the discussion section talking about multiple-item scales is a misplaced generalization. This study is about quick and repeated measures using a button pushing technique. Maybe if the subjects were given a paper and pencil exercise, you might not have uncovered any significant differences at all, using a simple dichotomy: "good" and "bad" classifications of some brands.

Some additional questions to ponder are:

- What is the effect of any accompanying information?

- What is the effect of availability and necessity for deliberation time?

- What is the effect of involvement on response time?

- What is the applicability for multiple exposure research in advertising?

DISCUSSION

Primarily, these three papers and similar research may be seen as something on-going in the area of understanding the effect of the various dimensions of information. If we assume that we are dealing with information which is relevant (salient and deterministic), accessible (either stimulus or memory based) and diagnostic, we can conceptualize several important dimensions of information. One such attempt is depicted in the Figure.

Whether we use a low/high degree of information, congruent/ incongruent type of information, or positive/negative information it will have an impact on the evoked set, which affects both judgment and final choice. However, an additional dimension which we have not successfully integrated in much of consumer research is the complexity introduced by the time dimension i.e. static versus dynamic process of information use.

The dynamism introduced by the time dimension impacts what information is presented when and in what manner for maximizing persuasibility either in advertising and in other sales contexts. For example, consumers may be "educated" to upgrade their purchase when they are shopping for houses or cars or most other durables by being presented information which is positive/negative and congruent/incongruent and try to stay within the range of the low/high continuum without causing any dysfunctionality due to information overload.

A related aspect important to policy makers is the regulation which may be required to force information disclosure to try to benefit consumers. How is prior information base assessed? To what degree can one reasonably hope to "educate" consumers, given a limited budget and a limited time frame? How much to "simplify" complex information? What information needs to be disclosed, how will it impact industry and what contributes to increasing consumer welfare are questions which need to be looked at in the context of "educating" consumers in social marketing contexts (Andreasen 1992).

REFERENCES

Alba, Joseph W. and J. Wesley Hutchinson (1987), "Dimensions of Consumer Expertise," Journal of Consumer Research, Vol. 13, 4 (March), 411-454.

Andreasen, Alan R. (1992), "A Social Marketing Consumer Research Agenda for the 1990s", Presidential Address delivered at the 1992 Vancouver ACR meeting.

Dickson, Peter R. and Alan G. Sawyer (1990), "The Price Knowledge and Search of Supermarket Shoppers," Journal of Marketing, Vol. 54, 3 (July), 42-53.

Puto, Christopher P. (1987), "The Framing of Buying Decisions," Journal of Consumer Research, Vol. 14, 3 (December) 301-314.

----------------------------------------